HPI-axis and heat shock protein (HSP) gene transcripts, and their responsiveness to stress in Atlantic salmon (Salmo salar L.) embryos and larvae

(1)

Norwegian College of Fishery Science

HPI-axis and heat shock protein (HSP) gene transcripts, and their responsiveness to stress in Atlantic salmon (Salmo salar L.) embryos and larvae

—

Renate Andersen

Master thesis in Aquamedicine (60 credits) November 2015

(2)

(3)

Acknowledgements

I

Acknowledgements

This master thesis was conducted at Nofima in collaboration with AquaGen and the Norwegian College of Fishery Science, UiT - The Arctic University of Norway.

I would like to express my gratitude to my supervisors at Nofima: Hanne Johnsen, Helge Tveiten and Lill-Heidi Johansen. Especially, thanks to Hanne for being so present and helpful the last couple of weeks, and to Helge for always having an answer to my questions. I would also like to thank Erik Burgerhout for helping me with absolutely everything, and Audny Johansen for assisting me in the lab and all the good talks. I have truly enjoyed my experience and time spent at Nofima. Thanks to my supervisor Ingvill Jensen at UiT for checking in on me from time to time.

Thanks to Maren Mommens and Nina Santi at AquaGen, for answering all my mails and for letting me visit AquaGen in Kyrksæterøra. Also, thanks to Morten Marienborg and Astrid- Elisabeth C. Hansen for all help and good talks at Kårvika.

Finally, I would like to thank my fellow students at the Fish health office, for all the good talks, laughter and support. Especially thanks to Mathias and Iris, who made this master period amazing after all. I would also like to thank Åshild, Iselin and Iris for keeping up with me at home, my friends for caring, and my loving parents for always answering my calls.

Renate Andersen, 22.11.2015

(4)

Sammendrag (Norwegian abstract)

II

Sammendrag (Norwegian abstract)

Stress kan føre til en betydelig innvirkning på fysiologien og helsen til individet senere i livet.

Under en produksjonssyklus av lakse-egg, blir eggene utsatt for ulike typer behandlinger, som sjokking og transport. Slike behandlinger ville i voksen fisk kunne utløse en stressrespons.

Stressresponser kan deles inn i primære, sekundære og tertiære responser. Den primære responsen består av to akser, hvor en av dem er hypotalamus-hypofyse-interrenal (HPI) aksen, som resulterer i en økning av sirkulerende kortikosteroider. På cellenivå, er heat-shock proteiner (HSP) en viktig del av den sekundære responsen. I embryo starter ikke syntesen av kortisol, som er den viktigste kortikosteroiden hos teleoster, før rundt klekking. Imidlertid har gener som er sentrale i HPI-aksen, og HSP-gener, blitt detektert i embryo i flere utviklingsstadier. Selv om HPI-aksen antagelig ikke er fullt utviklet før klekking, kan en stressor føre til endringer i genuttrykk.

Basert på dette, ble åtte sentrale gener fra HPI-aksen (crf1, crf2, pomcA1, pomcA2, pomcB, gr1, gr2 og mr) og to HSP gener (hsp70a og hsp90a4) hos atlantisk lakse- embryo utsatt for sjokking og transport, analysert for å se etter en eventuell behandlingseffekt. I tillegg til dette ble nylig klekkede larver, og larver ved startfôring som var utsatt for en stress test, analysert for å kunne beskrive ontogenien, og for å analysere mulige langtidseffekter av sjokking og transport. Relativt genuttrykk av hele-dyr ble analysert med bruk av revers transkriptase real time polymerase kjedereaksjon (RT-qPCR). Resultatene viste at genene var uttrykt i alle analyserte stadier gjennom utviklingen. HPI-akse genene viste en økning i relativt utrykk gjennom utviklingen, utenom gr1 og gr2 som viste en forsinket økning. HSP genene derimot viste et lavere uttrykk i larver ved startfôring enn i embryo og nylig klekkede larver. Relativt uttrykk av HPI-akse genene viste ingen spesifikke kortvarig eller langvarig forskjeller etter utsettelse for sjokking og/eller transport. HSP genene derimot, viste en akutt økning etter transport, men ingen langvarig effekter.

Resultatene fra dette studiet indikerer at sjokking og transport ikke er kraftige nok stressorer til at relativt uttrykket av HPI-akse genene blir forandret i embryo. Resultatene indikerer også at HSP genene mulig spiller en viktig rolle i den cellulære stressresponsen gjennom embryogenesen.

(5)

Abstract

III

Abstract

Exposure to stress may have a profound impact on the physiology and health of an individual later in life. During a production cycle of Atlantic salmon eggs, the eggs are subjected to different kind of handling, e.g. shocking and transport. Handling of this extent would have elicited stress responses in adult fish. Stress responses can broadly be divided into primary, secondary and tertiary response. The primary stress response consists of two pathways where one of them, the hypothalamus-pituitary-interrenal (HPI) axis, results in elevations of circulating corticosteroids. On a cellular level, heat shock proteins (HSP) play an important role as a secondary response. In embryos, cortisol, which is the main corticosteroid in teleosts, is not synthesized before around hatching. However, genes that are central in the HPI-axis and HSP genes have been detected in fish embryos at several developmental stages. Even though the HPI-axis is not fully developed a stressor may alter the gene expressions.

Based on this, eight genes central in the HPI-axis (crf1, crf2, pomcA1, pomcA2, pomcB, gr1, gr2 and mr) and two HSP genes (hsp70a and hsp90a4) were examined, in Atlantic salmon embryos subjected to shocking and transport. In addition, newly hatched larvae, and larvae at start feeding subjected to a stress challenge, were analysed to map the ontogeny of the genes, and to examine any possible long-term effects of the shocking and transport. Relative gene expression of whole-animal were analysed using reverse transcriptase real time polymerase chain reaction (RT-qPCR). The results showed that all genes were present in all samples examined throughout the development. The HPI-axis genes showed an increased relative expression level during development, except for gr1 and gr2 that showed a delayed increase probably due to maternal transfer. The HSP genes, however, had a low expression level at start feeding compared embryos and newly hatched larvae. The relative expression of the HPI- axis genes did not show any specific short-term or long-term differences in relative gene expression after exposure to shocking and/or transport. The HSP genes, however, showed an acute upregulation after transport, but no long-term effects.

The results of this study indicates that shocking and transport are not high enough stressors to alter the expression of the HPI-axis genes. They also indicate that the HSP genes may play an important role in cellular stress response during development.

(6)

Abstract

IV

1 Introduction

Atlantic salmon (Salmo salar L.) is intensively produced in fish farms in Norway, an industry that has had enormous growth during the last decade. This has led to an increased demand of salmon eggs, and numbers registered by the Norwegian directorate of fishery showed that 719 009 thousand eggs were transferred to hatcheries during 2014, an increase of 13 % from 2013 (Fiskeridirektoratet, 2015). When rearing salmonids, stressful events due to handling are unavoidable and include, among others, sorting, grading, transport, and shocking. Early developmental stages are sensitive and exposure to stress may have a profound impact on the physiology and health of an organism later in life (Groot, 1996; Tsalafouta et al., 2014). It has been shown that exposure to stressors during development results in permanent changes in stress coping phenotypes in mammals, birds, amphibians and fish (Tsalafouta et al., 2014).

1.1 Aquaculture and production of salmon eggs

A production cycle of salmon eggs and larvae normally starts at an egg production site, where eggs are reared until the eyed stage, after which they are transported to a hatchery (see Figure 1). The eggs hatch at the hatchery and are reared until they eventually become smolts, ready for transfer to seawater.

During salmonid development, there are periods where the embryos are more sensitive to external stimuli that need to be taken into account, to prevent increase in mortality. The first period where salmon eggs shows significant sensitivity to handling is between fertilization and the so called eyed-stage, i.e. the stage were the eyes show pigmentation (Egidius and Helland- Hansen, 1973; Groot, 1996). This is the period of early cell division, blastulation and epiboly, in which the embryo begins to take form (Gorodilov, 1996; Groot, 1996). During the eyed- stage, embryos are more robust, and this is the preferred stage where eggs can be handled without causing any harmful effects (Hayes, 1930; Groot, 1996). In production, two main handling events occur during this period of development; shocking of eggs and transport (Maren Mommens, AquaGen, pers.com). First, eggs are intentionally shocked by agitating the eggs enough to rupture the vitelline membrane surrounding the yolk in dead eggs, but not so much that normally developing eggs are damaged (Groot, 1996). This results in coagulation of the yolk proteins and causes dead eggs to turn white and opaque, and can therefore easily be

(10)

Introduction

8 sorted out (Groot, 1996; Carls et al., 2004). Simultaneously, eggs containing embryo with unusual small eyes are also sorted out (Maren Mommens, AquaGen, pers.com). The second handling event takes place around 375 day degrees (d°C) when eggs are transported to hatcheries. Transportation occurs on ice in special designed styrofoam boxes. Depending on the location of the hatchery, the transportation can take from several hours to a couple of days (Maren Mommens, AquaGen, pers.com). At the hatchery, eggs are transferred into hatching trays and only disturbed by removal of dead eggs. During this period, head and body regions are recognizable and the embryo can be seen to move freely within the chorion. Blood vessels grows out over the surface of the yolk, and the heart is actively pumping (Gorodilov, 1996; Groot, 1996). Normally, the hatching of yolk sac larvae takes place around 500d°C. The yolk sac larvae lie on the bottom of hatching trays on a substrate that supports them with keeping a desired upright position, until they swim up at the onset of exogenous feeding. The start feeding occurs approximately at 900d°C (Groot, 1996; Maren Mommens, AquaGen, pers.com). Physical disturbances encountered in aquaculture, such as shocking and transport, usually evokes a variety of responses in fish (Barton and Iwama, 1991).

Figure 1: Normal production of salmon eggs and larvae. The pictures are showing, from left; early cell division, epiboly, eyed-eggs, embryo right before hatching, and newly hatched larvae. The number in brackets indicates the day degrees.

(11)

Introduction

9 1.2 Stress in fish

Stress can be defined as a state where the dynamic homeostasis of an animal are threatened or disturbed by intern or extern stimuli, commonly termed stressors (Wendelaar Bonga, 1997). The stress response in vertebrates can broadly be divided into a primary, secondary and tertiary response as shown in Figure 2 (Iwama, 1998). The primary response includes neuroendocrine responses that results in measurable elevation of cortisol and adrenaline in the circulation (Sumpter, 1997). Thereafter, a secondary response is elicited, that includes cellular responses and changes in features related to metabolism, respiration, acid-base status, hydromineral balance and immune function (Mommsen et al., 1999; Gabriel, 2011).

Primary and secondary stress responses are adaptive if they result in a physiological response that allows a fish to maintain homeostasis (Donaldson et al., 2008). Prolonged stress can give rise to a tertiary response, which refers to aspects of whole-animal performance, such as changes in growth, reproduction, behaviour, resistance to disease and ultimately survival (Barton, 2002). In the present study, the focus will be on the primary and secondary responses.

A wide range of stressors elicits both the neuroendocrine and cellular stress responses (Ackerman et al., 2000).

Abiotic and biotic stressors

Primary stress response Neuroendocrine responses -> Catecholamines↑

-> Corticosteroids↑

Secondary stress response Cellular response

-> Heat shock proteins↑

Metabolic changes

Osmoregulatory disturbances Changes in immune functions Hydromineral balance

Tertiary stress response

Changes in whole-animal characteristics:

- growth - reproduction - behaviour

- resistanse to disease -> survival

Figure 2: A figure showing the three grouped stress responses and their action (Modified from Barton et al., 2002).

(12)

Introduction

10 1.2.1 Neuroendocrine stress response

A stressor activates a two component system in fish; the hypothalamus-sympathetic- chromaffin (HSC) axis and the hypothalamus-pituitary-interrenal (HPI) axis (Wendelaar Bonga, 1997).

The HSC-axis leads to an adrenergic response. When elicited, sympathetic nerve fibres stimulates chromaffin cells in the head kidney to release catecholamine hormones, adrenaline and noradrenaline, into the circulation (Sumpter, 1997; Wendelaar Bonga, 1997; Reid et al., 1998). Catecholamines, predominantly adrenaline in teleosts, are both synthesized and stored in the chromaffin cells and can therefore be rapidly released after stress (Reid et al., 1998;

Barton, 2002). One of the primary roles of plasma catecholamines is to modulate cardiovascular and respiratory functions in order to maintain adequate levels of oxygen in the blood. In addition, they serve to mobilize energy stores to provide for the increased energy demands that often are required after exposure to stressors (Reid et al., 1998).

The HPI-axis consists of a three stage endocrine pathway as shown in Figure 3, where cortisol is the physiologically important hormone responsible for the effect of stress (Sumpter, 1997;

Mommsen et al., 1999). A hormone cascade is initiated by external stimuli that stimulate the hypothalamus to release corticotropin-releasing factor (CRF). CRF will further stimulate corticotrophin cells in the anterior pituitary to secrete adrenocorticotropic hormone (ACTH) into the circulation, which thereafter stimulates the interrenal cells, in the head kidney, to produce and secrete corticosteroids, mainly cortisol in teleost fishes (Sumpter, 1997;

Wendelaar Bonga, 1997; Mommsen et al., 1999; Flik et al., 2006). The release of cortisol is delayed relative to catecholamine release (Reid et al., 1998; Barton, 2002). The main role of cortisol is to meet the energy demands in a stress response, by redirecting the metabolism, and to limit the defence reactions to stress in order to protect the body from further damage (Xiong and Zhang, 2013; Wendelaar Bonga,1997). In the present study, the focus of neuroendocrine stress response will be on the HPI-axis, and the following sections will go into more detail about some of its important components.

(13)

Introduction

11 1.2.1.1 Cortisol releasing factor (CRF)

Cortisol releasing factor (CRF) is a neuropeptide that is produced in nucleus preopticus (NPO) of the hypothalamus. Various stressors are associated with an increased expression of preoptic area CRF, in adult fish (Bernier and Bristow, 2008). In teleosts, CRF controls the HPI- axis through activation of specific G-protein coupled CRF receptors (CRF-R1 and CRF-R2) and is regulated by a shared CRF binding protein (CRF-BP). In addition to the regulation of the endocrine stress response, other functions of CRF include for example food intake inhibition and behavioural modulation (Alderman and Bernier, 2009). In fishes, CRF is also produced and secreted from the caudal neurosecretory system (CNSS), a unique organ located at the caudal end of the spinal cord. At the hypothalamic level, CRF is considered to be the major regulator of adrenocorticotropic hormone (ACTH) secretion from the pituitary and thereby plays a key

Figure 3: The HPI-axis and the hormone cascade leading to cortisol secretion. CRF is projected directly from the hypothalamus to the pituitary where it binds to CRF-receptors in corticotropic cells. The binding eventually elicit a secretion of ACTH into the bloodstream. POMC is the precursor for ACTH. Circulation ACTH binds to MC2-receptors in interrenal cells of the head kidney and stimulates to synthesis and eventually secretion of cortisol into the bloodstream. Cortisol enters the target tissues/organ by diffusion and binds to GR and MR, which mediates the action of cortisol by altering target gene expression. An elevated level of cortisol has a negative feedback on the hypothalamus and pituitary.

(14)

Introduction

12 role in coordinating the neuroendocrine, autonomic, and behavioural responses to stress (Alsop and Aluru, 2011).

1.2.1.2 ACTH/POMC

Binding of CRF to CRF R1/R2 in corticotropic cells of the anterior pituitary, elicits the release of ACTH into the circulation. ACTH is derived from a precursor hormone termed proopiomelanocortin (POMC), which is a large polypeptide that is progressively cleaved by prohormone convertases into several biologically active peptides (Flik et al. 2006; Nelson and Cox, 2008). In humans the POMC gene expression is stimulated by corticotrophin-releasing hormone (CRH) and vasopressin, and is suppressed by glucocorticoids (Raffin-Sanson and Bertagna, 2003). The peptidescan be broadly divided into three groups: adrenocorticotropic hormone (ACTH)-like, endorphin-like and MSH-like products. POMC is primarily synthesised in two cell types of the pituitary gland: the corticotrophs of the anterior lobe and the melanotrophs of the intermediate lobe, each lobe being responsible for different peptide products (Sumpter et al.1997; Mosconi et al., 2006). In brown trout, handling and confinement has only showed to activated the corticothrophs, whereas when the handling was combined with thermal shock, both corticothrophs and melanotrophs were activated (Sumpter et al.

1985). POMC has, in addition, shown to be present in a variety of other brain regions, and peripheral tissues such as the skin (Hansen et al.2003; Karsi et al. 2004). ACTH is recognized as the principle stimulator of cortisol release (Wendelaar Bonga, 1997; Flik et al.2006).

Circulating ACTH binds tomelanocortin 2 receptor (MC2R) in the steroidogenic interrenal cells embedded in the head kidney in teleosts. The binding to MC2R stimulates adenylate cyclase and cAMP-dependent signalling pathways, to stimulate cortisol synthesis. This receptor has shown to be downregulated following stress due to negative feedback control (Alsop and Aluru, 2011).

1.2.1.3 Cortisol and the cortisol receptors

Free circulating cortisol enters target cells, such as hepatocytes, by passive diffusion, where its action is mediated by the glucocorticoid receptor (GR) and mineralocorticoid receptor (MR) (Alsop and Aluru, 2011). These receptors belong to the nuclear receptor superfamily of ligand- bound transcription factors. The receptors require the presence of certain heat shock proteins (e.g. HSP70 and HSP90; see section 1.2.2) to form a steroid compatible heterocomplex (Norris

(15)

Introduction

13 and Hobbs, 2006). When activated the receptors are translocated into the nucleus of the cell, where they acts as transcription factors involved in the activation or silencing of specific genes (Li and Leatherland, 2012). In fishes, GR and MR are expressed in a variety of tissues including liver, gill, muscle, kidney, blood, and brain (Norris and Hobbs, 2006). Mediated by GR, cortisol modulate aspects of metabolism, growth, reproduction and immune function during a stress elicited response (Wendelaar Bonga, 1997). The roles of MR and its ligand are less clear. The main ligand to MR in mammals is aldosterone, due to inactivation of cortisol by an enzyme, 11_HSD2, allowing aldosterone to bind. Teleosts lack the capacity to synthesize aldosterone, but another possible MR ligand that is studied in rainbow trout is 11 deoxycorticosterone (DOC) (Sturm et al.2005). Cortisol elicits a negative feedback primarily at the brain to repress the release of CRF, thereby reducing ACTH secretion and, ultimately, cortisol secretion (Norris and Hobbs, 2006; Bumaschny et al., 2007; Alsop and Aluru, 2011)

1.2.2 Cellular stress response

Cortisosteroids along with catecholamines, mediate secondary stress responses, in which a cellular response is one of them (Donaldson et al., 2008). Fish, like other organisms, produce a variety of proteins as part of the stress response, which are included in a generalized response system that exists at a cellular level. These proteins, which commonly are called stress proteins include among others, metallothioneins and heat shock proteins (Wendelaar Bonga, 1997; Basu et al., 2002). The heat shock proteins (HSP) are one of the most common and most studied groups of stress proteins in the cellular response (Deane and Woo, 2011).

1.2.2.1 Heat shock proteins

Heat shock proteins (HSPs) are a group of highly conserved intracellular proteins first detected in fruit fly when exposed to heat shock. They are classified into families based on their protein molecular size (kDa) which also gives them their names: HSP100, HSP90, HSP70, HSP60, and the small HSPs (Deane and Woo, 2011). HSPs are expressed in all tissues and cells constitutively, but some are also inducible in response to biotic or abiotic stressors. In an unstressed cell, the constitutive HSPs generally function as molecular chaperones assisting the folding of nascent polypeptides, protein folding, translocation of proteins, and degradation of misfolded proteins (Basu et al., 2002; Deane and Woo, 2011). When exposed to a stressor the inducible HSPs are upregulated, which in turn gives the cell added protection to repair and

(16)

Introduction

14 prevent damage from cellular stress associated with protein denaturation (Iwama et al., 1998;

Basu et al., 2002; Donaldson et al., 2008; Deane and Woo, 2011). Most of the inducible HSP genes do not contain introns and therefore their mRNA are rapidly translated into protein within minutes after an exposure to a stressor (Morimoto et al., 1992; Iwama et al., 1998).

HSPs have been found to be upregulated when subjected to both high and low temperatures, and it is also widely accepted that their expression can alter upon exposure to a range of other abiotic, as well as biotic and chemical stressors (Deane and Woo, 2011; Donaldson et al., 2008). HSPs are also known to play key roles during embryonic development (Deane and Woo, 2011). A number of HSPs are expressed at high levels during normal cell growth and has shown to be important for reducing temperature-induced damage and deformities of fish embryos (Iwama et al., 1999; Donaldson et al., 2008).

The HSP70 family represents the most abundant and the most highly conserved HSPs. HSP70 is composed of constitutive (HSC70) and stress-inducible (HSP70) isoforms. Inducible isoforms are the best studied HSP70 in developing zebrafish and are frequently induced by thermal stress (Rupik et al., 2011). The constitutive members play important chaperoning roles in unstressed cells (Basu et al.2002). In addition, it has been shown in zebrafish that Hsp70s are required during the normal process of lens development under non-stress conditions (Evans et al. 2005).

Members of the eukaryotic hsp90 family interacts with and modulates the activity of important cellular signalling molecules, such as steroid receptors and transcription factors (Krone et al., 2003). It has been estimated that HSP90 accounts for about 1% of the total soluble protein in the cytosol of an unstressed cells, which makes it one of the most abundant proteins (Picard, 2002). Vertebrates express two hsp90 genes, hsp90α and hsp90β, and studies in zebrafish, indicate that these genes are differentially regulated (Basu, 2002; Krone et al.2003). HSP90 interacts with a large number of proteins, and its interaction with steroid receptors, including GR, results in the formation of a stable heterocomplex (Pratt and Toft, 1997). The binding of HSP90 increases the receptor stability by allowing GR to be conformational competent for ligand binding, in addition to prevent proteasomal degradation of GR. HSP90 is thereby a key molecular chaperone that is crucial for cortisol mediated cellular action, including GR signal transduction (Pratt and Toft, 1997). In addition, the isoform HSP90α

(17)

Introduction

15 has been shown to be required for normal muscle development in zebrafish during embryogenesis (Krone et al., 2003).

1.3 The genes of the HPI-axis

In the evolution of vertebrates, several whole genome duplication (WGD) events are thought to have occurred. One of them occurred 320–350 million years ago specifically in an ancient fish, which gave rise to a number of duplicate genes that exist exclusively in teleost today (Alsop and Vijayan, 2009). In addition, salmonids have gone through another WGD, which occurred about 25-100 million years ago in a common ancestor (Meyer and Schartl, 1999).

After a WGD, most duplicate genes return to single gene systems, but in some instances, duplicate genes are retained (Alsop and Vijayan, 2009).

Two CRF genes have been found in several fish species, including white sucker; Catostomus Commersoni, carp; Cyprinus carpio, sockeye salmon; Oncorhynchus nerka, and rainbow trout;

Oncorhynchus mykiss (reviewed in Alsop and Vijayan, 2009), while only a single CRF system is reported in zebrafish; Danio rerio (Chandrasekar et al., 2007; Alsop and Vijayan, 2008).

According to Alsop and Vijayan, (2009) the duplicate CRF sequences are so similar that only a few studies have been able to differentiate the two. Doyon et al. (2003) was able to differentiate between the CRF paralogs in rainbow trout brain, which showed that levels of both CRFs were highest in the preoptic area of the hypothalamus, and were expressed to the same extent.

Similar to the CRF genes, most POMC genes are identified in duplicates among studied species, including common carp; Cyprinus carpio (Arends et al., 1998), zebrafish (Nunez and Gonzalez- Sarmiento, 2003) and sockeye salmon (Okuta et al., 1996). In rainbow trout two POMC genes, in addition to a splice variant of one of the genes, has been identified (Salbert et al., 1992;

Leder and Silverstein, 2006).

Two GR genes and one MR have been found in several teleost, except for zebrafish which only have identified a single GR, a homolog to GR2 in other teleosts (Alsop and Vijayan, 2009).

Between the two GRs, GR2 has shown to be more sensitive to cortisol than GR1 in rainbow trout (Bury et al., 2003). As mentioned in section 1.2.1, deoxycorticosterone (DOC) is studied

(18)

Introduction

16 in rainbow trout to be a possible ligand for MR (Sturm et al. 2005). The presence of three receptors with different affinities for cortisol raises, according to Norris and Hobbs, (2006), a possibility that the response to cortisol in tissues may change, as increasing levels of cortisol receptors with lower affinities are bound, and activated at higher cortisol levels.

1.4 Ontogeny of HPI-axis hormones in addition to HSP70 and HPS90

In fish, the most commonly used stress indicator is, to the author’s knowledge, the elevation of cortisol (Gabriel, 2011). In fertilized eggs, embryos and larvae, changes in cortisol content at various developmental stages have been examined in several fish species (reviewed by Pittman et al. 2013). In fertilized eggs, the cortisol content is of maternal origin and, according to Pittman et al. (2013), seems to be necessary for the metabolic needs and for the development of various organs during early development. Most of the examined species mentioned show a general pattern of changes in cortisol content in the egg, with relatively high levels after fertilization, followed by a decrease throughout embryogenesis as the maternal deposited cortisol is depleted. The lowest levels are registered around the time of hatching and, thereafter, the larva begins to synthesize cortisol and basal levels increase (Alsop and Aluru, 2011; Pittman et al., 2013). The timing of de novo synthesis of cortisol varies among species. Studies on rainbow trout indicates a cortisol synthesis 6 days before hatching (Auperin and Geslin, 2008). In both chinook salmon; Onchorynchus tshawytscha and zebrafish de novo synthesis of cortisol was detected around the time of hatching (Feist and Schreck, 2001; Alsop and Vijayan, 2008). From early embryogenesis, expression of other HPI-axis genes also has been shown to undergo dynamic changes, suggesting that they are functional at this time (Alsop and Aluru, 2011). Most studies have been conducted on zebrafish, and therefore, the following description of the ontogeny of different HPI-axis genes will mainly be based on findings in zebrafish embryos and larvae.

Crf has been detected in zebrafish throughout embryogenesis and larvae showed an increase in crf expression levels between hatching and exogenous feeding (Alderman and Bernier, 2009). Crf transcripts have been detected during embryogenesis for several other teleost species including tilapia; Oreochromis mossambicus (Pepels and Balm, 2004), rainbow trout (Fuzzen et al., 2011), and European sea bass, Dicentrarchus labrax (Tsalafouta et al., 2014). A study on rainbow trout showed a pattern of crf expression throughout ontogeny that was

(19)

Introduction

17 similar to cortisol, and the larvae showed a peak in CRF mRNA levels occurring at 56 days post fertilization (dpf); the onset of exogenous feeding (Fuzzen et al., 2011).

Hansen et al. (2003) registered pomc mRNA expression in fertilized zebrafish eggs by RT-PCR, which almost completely disappeared within the next few hours after fertilization, which demonstrates maternal expression. After 18hpf, zygotic pomc RNA synthesis levels significantly increased with a maximum at 28 hpf (Hansen et al., 2003).

The expression pattern of gr and mr has shown to be distinct during the embryogenesis in zebrafish. Expression levels of mr showed a continuous elevation during development from fertilization until start feeding, while those of gr followed closely the cortisol profiles seen in the embryos; i.e. showing a decrease throughout embryogenesis, followed by a rise around hatching, which continued until start feeding (Alsop and Vijayan, 2008)

HSPs studied in zebrafish, have shown to be expressed in spatial and temporal patterns, which coincided with the assumed targets of their chaperoning activity (Krone et al. 1997). It was also shown that several HSPs may be directly involved in embryonic cellular differentiation (Martin et al.2002). Both constitutive and inducible forms of hsp70 have been detected in the developing zebrafish. During embryogenesis of zebrafish, basal levels of inducible hsp70 showed to be low, while constitutive members of hsp70 have shown to be more strongly expressed (Lele et al., 1997; Santacruz et al., 1997). Embryos of Atlantic salmon expressed hsp70 mRNA transcripts at all examined stages from 62d°C until 200d°C (Takle et al., 2005).

Hsp70 expression was upregulated in early larval stage of zebrafish, which also was registered in silver sea bream (Sparus sarba) larvae after 14 dph (Yeh and Hsu, 2000; Deane and Woo, 2003). Deane and Woo, (2003) registered that hsp90 increased from 1dph and onwards, where the profiles during 1-14 dph was parallel to cortisol. In addition, the two isoforms, hsp90α and hsp90ẞ, have shown to be expressed in zebrafish during embryogenesis (Krone et al.1997).

1.5 Stress response in early development

Several studies have registered that the necessary components for a functioning HPI axis are in place before, or at the time of hatching. However, there has not shown to be any stress-

(20)

Introduction

18 induced elevation of cortisol during early embryogenesis (reviewed by Pittman et al.2013). In fact, stress-induced cortisol alterations have, to the author’s knowledge, not been detected before hatch in most species studied. In rainbow trout, stress-induced cortisol elevation was not detected before 11 dpf (Auperin and Geslin, 2008) or 14 dpf (Barry et al., 1995a). In chinook salmon stress induced elevation of cortisol was detected one week after hatching (Feist and Schreck, 2001). In rats, a 2-week stress hyporesponsive period has been shown where stressors do not elicit an increase in circulating glucocorticoids levels, as they do in adult animals. This period is thought to be a critical time where corticosteroids may have permanent effects on the neural organization and development (Barry et al., 1995a; Barry et al., 1995b).

Early development represents a critical period during life history of fishes (Groot, 1996). At this time of the development environment may irreversibly influence the phenotype (e.g.

morphology, physiology, behaviour) by allowing rapid adaptations. These adaptations may be beneficially for the animal later in life, or in contrast give adverse consequences if there is any mismatch between the anticipated and the actual environment later in life (Pittman et al., 2013). The hypothalamus-pituitary-adrenal (HPA) axis in mammals is highly susceptible to

‘programming’ during the development (Xiong and Zhang, 2013).

Thus, even though the HPI-axis is not fully functional before hatching in most teleosts studied, the genes of the HPI-axis may still be altered by stressors, which in turn might affect the individual later in life. Auperin and Geslin (2008) detected in rainbow trout that stress applied to eyed-egg, eggs at the time of hatching, and yolk sac larvae, resulted in a reduced cortisol response to stress in fingerlings. They suggested that stress during the development of the HPI-axis may have long lasting effects and may influence the fish’s ability to cope with stress later in life.

(21)

Introduction

19 1.6 Aim

This master project was part of a project conducted by Nofima Tromsø in collaboration with AquaGen Norway. It was important to clarify several aspects related to some of the possible effects mechanical shocking and transportation of eggs may have. The main objective of this study was to examine gene expressions of HPI-axis - and HSP genes in Atlantic salmon (Salmo salar L.) embryos and larvae in terms of upregulation or downregulation after exposure to shocking and/or transport by analysing total RNA from whole eggs and larvae with RT-qPCR.

Sub goal 1:

Examine the ontogeny of eight HPI-axis genes (crf1, crf2, pomcA1, pomcA2, pomcB, gr1, gr2 and mr) and two HSP genes (hsp70a and hsp90a4) by using four time points in the embryo and larval development. Examine if shocking and transport conducted in rearing of salmonid eggs (mechanical shock and transportation on ice), results in alteration of the relative expression of these genes.

Sub goal 2:

Examine if the shocking and transport alters the expression of the mentioned HPI-axis genes and HSP genes over time, by examining the relative gene expression in larvae at start feeding.

In addition, examine if the shocking and transport may lead to different gene expression levels when the larvae at start feeding are exposed to a stress challenge.

(22)

Materials and Methods

20

2 Materials and Methods

All chemicals and kits used are listed in appendix I.

2.1 Eggs, fertilization and incubation conditions

Atlantic salmon (Salmo salar) eggs from one female and milt from one male were sent by plane from AquaGen, Kyrkseterøra to Nofima, Tromsø. The package containing eggs and milt was immediately transported by car for 30 minutes to the Aquaculture Research Station in Kårvika. The eggs and milt were during the whole transport kept in a Styrofoam box; packed in plastic bags and laid between two layers of ice covered with newspaper. The eggs and milt held a temperature of 2-3°C at the time of unpacking. The whole transportation took less than 24 hours.

2.1.1 Fertilization

Approximately 1,825 L eggs were fertilized immediately after transportation to Kårvika, using a dry fertilization method obtained from AquaGen. The eggs were carefully poured into a tub where they were washed with a washing solution (recipe, AquaGen) until blood and ovary fluid was removed. The beam was always pointed towards one of the sides of the tub, avoiding directly contact with the eggs. After the washing procedure, the tub was filled with washing solution equivalent to 1/3 of the egg volume. Approximately 2 mL milt per litre of eggs was added. The milt and the eggs were then evenly distributedwith gently stirring. After 25 seconds, activation solution (recipe, AquaGen) was added to activate the milt. The mixture was carefully stirred and left for 2,5 minutes. The milt and activation solution were washed away using the washing solution. Directly after the fertilization the eggs were treated with a disinfection solution composed of 10 parts Buffodine (Evans Vanodine) to 1000 parts water (50mL Buffodine + 5L water from the hatching column). The disinfection solution was poured into the tub covering the eggs. It was left for 10 min and then gently washed away with the washing solution. The eggs were finally divided into beakers each containing approximately 175 dL eggs, and transferred to 12 plastic incubator boxes in a hatching column (see chapter 2.2.2) for swelling.

(23)

21 2.1.2 Incubation

Incubation of eggs and embryos was carried out in a specially designed hatching column with temperature regulation and continuous water supply that was filtered and free of chloride. The hatching column held three incubation units, A, B and C, each containing four special incubation boxes as shown in Figure 1. The boxes were labelled with four groups, each group having triplicates that were evenly distributed in the column to minimize ‘’tank effect’’ (see Figure 4). Over the first nine days after fertilization, the

water temperature was slowly increased from 3°C to 7°C to acclimate the eggs. During the rest of the experimental period the eggs and larvae were held under a mean (±SD) water temperature of 7,04 (± 0,16) °C and dissolved oxygen levels of 101,7 (± 1,53) %. Opaque, white eggs were counted and eliminated once a day. Both the incubation units and the whole hatching column was covered with opaque plastic sheets at all time to avoid light entering the incubation boxes.

Figure 5: A flow chart showing the hatching column containing the units (A, B and C) and the incubation boxes divided into four groups of triplicates.

2.2 Experimental design

Eggs were divided into four groups of triplicates, with each group going through different treatments. Group 1 was the control group, group 2 was shocked and transported, group 3 was shocked and group 4 was transported (see figure 5). At the end of the experimental period, all groups were subjected to a stress challenge test to evaluate the effect of the previous treatments.

Figure 1:

Figure 4: A picture showing one of the incubation units containing four boxes with salmon eggs.

(24)

22 2.2.1 Shocking of eggs

Eggs from groups 2 and 3 were shocked at around 326 day degrees, one incubation box at a time. Eggs from an incubation box were gently poured into a bucket filled with 7,5cm water.

The eggs were then poured into another bucket, containing the same amount of water, from a height of 60 cm over the water surface as shown in Figure 6. This procedure was repeated three times for all replicates within the two groups.

Figure 6: A picture showing shocking of salmon eggs as described in chapter 2.3.1.

2.2.2 Shipment on ice

Eggs from groups 2 and 4 were transported when the eggs were 377d°C. They were transferred to a Styrofoam box specially designed for transportation of fish eggs (obtained from AquaGen). The box contained three shelves, each of them divided into 12 units keeping the eggs separate. All the shelves contained small holes in the bottom to enable water to run through. During the transport, the upper shelf was filled with ice so that water could drain through to the next shelf containing eggs. It was important that the eggs were wet during the whole transportation but not soaked in water. The transportation lasted for 48 hours. The temperature in the box was measured at all times by two gauges that were placed separately in two empty units on the same shelf as the eggs. During the transportation the two gauges showed mean (±SD) temperature of 1,12 (± 0,96) and 1,41 (± 0,91)°C.

2.2.3 Stress test

Larvae from all groups were stressed when they were approximately 918d°C. Approximately 20 larvae from an incubation box were continuously transferred as described in section 2.3.4, and directly distributed into four special cylindrical incubation devices where three of the cylinders were stress challenged. The latter cylindrical incubator was immediately euthanatized and sampled as control. The stress challenge test was performed by exposing

(25)

23 larvae acclimated to 7°C, to ice water for 1 min and thereafter to air (14°C) for 1 min. After the challenge, the cylinders were placed in suitable containers placed in the hatching column so that the stressed larvae did not get disturbed between the challenge and the time of sampling.

2.2.4 Sampling

Samplings were conducted at four time points as shown in Figure 7. Three of the time points were sampled in accordance with the different treatments (described in section 2.3.1 – 2.3.3);

before the treatment started, and 1 hour, 3 hours and approximately 24 hours after the treatment was finished. The fourth sample was taken a few days after hatching. Two types of samples were taken at all time points; one in RNA-later® (Ambion) for gene expression analyses and one immediately frozen in Liquid N2 for other analyses. The RNA-later samples contained maximum 20 eggs or larvae and approximately 10 mL of RNA-later® (Ambion) solution in 20 mL tubes. Following samplings the tubes containing RNA-later® (Ambion), were stored as recommended by the manufacturer; overnight at 4 °C and then frozen at -20 °C the following day. The samples that were immediately frozen in liquid N2, contained a maximum of 5 eggs or larvae in 1,8 mL Cryo Tubes (Nunc). The samples were transferred to a -80 °C freezer the same day as the sampling.

Figure 7: A flow chart containing an overview over the rearing period (blue arrows) and the time of treatments (yellow/orange triangles) performed on the four different groups. The orange arrows indicates the time of sampling.

(26)

24 Sampling of eggs:

A special egg tweezer or a spoon was used to collect random eggs on a petri dish kept on ice while the sampling was conducted. Normal eggs were sampled on RNA-later® (Ambion) and liquid N2. Eggs that were developing at a slower rate (see figure 8) were sampled on separate tubes. Underdeveloped, blank eggs were discarded (see figure 8).

Sampling of larvae:

Larvae were collected using a transparent tube and the siphon principle, which made it possible to collect an almost exact sample size without affecting the remaining larvae. All larvae that were sampled were euthanatized using an overdose of Benzocaine. They were laid on a petri dish on ice and, in the same way as the sampling off eggs; normal larvae were selected and transferred to tubes containing RNA-later® (Ambion), and tubes that were put on liquid N2. Larvae that were un-normal (see figure 9) were collected in a second tube containing RNA-later ® (Ambion).

Normal Small-eyed Blank

Un-normal larvae Normal larvae

Figure 8: Pictures showing different developed eggs; normal, small- eye and blank (stopped developing).

Figure 9: Picture showing normal and un-normal larvae

(27)

25 2.3 Quantitative RT-PCR

Quantitative real-time polymerase chain reaction (RT-qPCR) has become the leading tool for the detection and quantification of DNA or RNA (as cDNA). Using sequence-specific primers, the number of copies of a particular DNA or RNA sequence can be determined. The RT-qPCR measures the amount of DNA or cDNA after each cycle and thereby monitors the progress of the PCR reaction as it occurs in real time. Fluorescent dyes are used as dictation agents where the yield of increasing fluorescent signal are in direct proportion to the number of PCR product molecules (amplicons) generated. One fluorescent dye that is suitable to use is SYBR Green.

SYBR green is a fluorescent DNA-binding dye that binds to any double stranded (ds)DNA and provides a fluorescent signal that reflects the amount of DNA product in the sample (LifeTechnologies, 2011b, LifeTechnologies, 2014).

Figure 10: A flowchart that shows the main steps of quantitative RT-PCR.

In this experiment RT-qPCR was used to examine gene expression from HPI-axis genes of corticotropin releasing factor (CRF), proopiomelanocorticoid A1, A2 and B (POMC A1, POMC A2 and POMC B), glucocorticoid receptor 1 and 2 (GR 1 & GR 2), mineralocorticoid receptor (MR) and heat shock protein 70a and 90a4 (HSP70a and HSP90a4). Eukaryotic Elongation factor 1 alfa (ef-1-α), Ribosomal 18S RNA (18S rRNA) and Beta-actin (ẞ-actin) were used as housekeeping genes. Primers were designed using the Primer Express 3 software (Life Technologies) and synthesized by Eurogentec. All primers are listed in Appendix II. Ten individuals from each group were analysed, which equals five individuals from two of the replicates (incubation unit B and C).

2.3.1 RNA isolation

RNA isolation was conducted in three steps; homogenization, purification of nuclei acids and DNase treatment. Whole eggs and larvae were used. One random egg from each group, from each sampling date, was weighed before the homogenization. This was done because the

RNA isolation cDNA synthesis qRT-PCR

(28)

26 weight of the sample tissue needed to be known for further purification. RNase-free equipment was used during the whole procedure.

2.3.1.1 Homogenization

To homogenize salmon eggs and larvae MagMAX-96 Total RNA Isolation Kit (Ambion) was used and the procedure was conducted according to the manufacturers protocol, with some minor modifications. The amounts needed to homogenize one salmon egg without making the homogenate too viscous had earlier been established in the lab. Salmon eggs on RNA- later® (Ambion) were thawed on the bench, punctured with a pipette tip, and put into tubes containing ceramic beads (Precellys) and 800 µL of Lysis/binding solution concentrate (Ambion). The samples were homogenized using the machine Precellys 24 lysis and homogenization (Bertin technologies) for 3 x (15 seconds x 6800rpm), with 30 seconds pause between each round. The tubes containing the homogenate were cooled for about 10 minutes and then added 20 µl of Proteinase K (Ambion). After incubation at 37 °C for 90 minutes the homogenate was frozen at -80 °C until further analyses.

2.3.1.2 Total nucleic acid isolation

The MagMax^TM- 96 Total Kit from Ambion was used to extract RNA from salmon eggs. The procedure was conducted according to the producers protocol, except from the DNase treatment which was done in a separate step (see chapter 2.4.2.1). Homogenized samples were thawed, mixed and centrifuged for 2min at 2500rpm (Kubota 1300). All solutions and plates were prepared as described in the manufacturers protocol. In brief, the reaction volumes (180µL) of the test-plate contained 5mg homogenate adjusted to 100 µL with Lysis/binding solution concentrate (Ambion), 20 µL Bead mix solution and 60 µL 100 % isopropanol Prima (Arcus). In addition to the test-plate, plates containing washing solutions, elution buffer and special tip compounds were prepared. Total nucleic acids were isolated using a magnetic purification machine; MagMAX^TM Express 96 (Applied Biosystems). The machine purifies the samples by magnetically capturing the RNA binding beads in the homogenate and washing them in several steps to remove cell residues, proteins and other contaminants (AppliedBiosystems, 2008, LifeTechnologies, 2011a). A MagMax Express plate (Applied Biosystems) containing the eluate with the purified nucleic acids was put on an Ambion magnetic-ring stand (Applied Biosystems), thereby making it possible to collect the

(29)

27 eluate without getting remains of the magnetic beads. The now isolated total nucleic acid samples were held on ice or frozen at -80 °C until further analysed.

2.3.1.3 DNase treatment

To clean the RNA from contamination of genomic DNA the nucleic acid sample TURBO DNAfree Kit (Ambion) was used in accordance to the manufacturers protocol (LifeTechnologies, 2012). Centrifugation was conducted in a Jouan A14 centrifuge at 10000xg in 2min. After the procedure, the supernatant containing the isolated RNA was collected and frozen at -80 °C until analysed further.

2.3.2 NanoDrop

NanoDrop (Saveen Werner AB) is a spectrophotometer using fibre optic technology and surface tension to hold 0,5-2 µL of sample in place between two optical surfaces.

(ThermoFisher, 2015). NanoDrop determines the RNA concentration by measuring its absorbance at 260nm (A260). Additionally it measures the purity of the RNA sample which mainly are shown by the A260/A280 ratio (LifeTechnologies, 2011a). All the isolated RNA samples were measured on a NanoDrop 8000 (Thermo scientific) before cDNA synthesis, to assess RNA quality and quantity. Elution buffer (Ambion) was used as blank.

2.3.3 cDNA synthesis

RNA is not suitable as target for DNA polymerase and must be reversely transcribed to complementary DNA (cDNA) before it can be analysed with RT-qPCR. To reverse transcribe the isolated total RNA, High-Capacity cDNA reverse Transcription Kit (Applied Biosystems) was used in accordance with the manufacturer’s protocol. The isolated RNA samples were thawed on ice and heated at 60 °C for 5 minutes to minimize secondary structures in the RNA. In brief, reaction volumes of 25 µl contained 200ng RNA, 2.5 µl 10x Reverse Transcription buffer, 1 µl 25x dNTPs, 2.5 µl 10x Random Primer, 1 µl Oligo d(T), 1.25 µl Multiscribe Reverse Transcriptase and 1.75 µl Nuclease free water (Ambion). The reaction was done in 96-well plates (Bioplastics).The plate was carefully mixed, briefly centrifuged and placed in the PCR machine 2720 Thermal cycler (Applied Biosystems) using the following cycle parameters; denaturation for 10 minutes at 25 °C, annealing for 120 minutes at 37 °C and elongation at 85 °C for 5minutes before the temperature decreased to 4 °C . The newly synthetized cDNA was diluted

(30)

28 1:8 in nuclease-free water (Ambion) and used as a stock solution. An aliquot of the stock- solution was additionally diluted 1:40 for use in the RT-qPCR analyses. The stock-solutions and aliquoted user-solutions were stored at -20 °C until further use.

2.3.4 RT-qPCR

Quantitative RT-PCR was used to study relative differences in gene expressions of central stress related genes. RT-qPCR assays of the HPI-axis genes and HSP genes were established by Hanne Johnsen. All primer pairs gave single distinctive melting peaks verifying the absence of primer dimers and other unwanted amplification products. The amplification efficiency of each primer pair were calculated using a 2-fold dilution series with 11 dilutions, starting with cDNA diluted 1:10 from Larvae at 900 day degrees in agreement with the following equation:

Primer efficiency (E) was calculated following the equation 𝐸 = 10(−1/𝑠𝑙𝑜𝑝𝑙𝑒) (Pfaffl, 2001) Quantitative RT-PCR was conducted using the 7900HT Fast Real-Time PCR system (Applied biosystems). The RT-qPCR was run in duplicates with each reaction containing 10µL Power SYBR Green Master Mix (Life Technologies), 1.2 µL (300 nM) of each primer, 0.6 µL nuclease free water (Ambion) and 7µL diluted cDNA to a final concentration of 20 µL. Two different controls were included in the RT-qPCR setup for each primer pair and plate; a no template control using nuclease free water (Ambion) as template instead of cDNA, and a positive control where a pre-made standard pool of cDNA was used as template. Additionally, random DNase treated RNA samples were used as templates to test for possible genomic contamination. When ready, the 384-well plate (Applied Biosystems) was covered with MicroAmp Optical Adhesive Film (Applied Biosystems) and briefly centrifuged in a Jouan RC 10.22. A template-file was made using the SDS 2.3 software (Applied Biosystems), and a PCR- program with the following cycling parameters were initiated: denaturation at 90 °C for 10 minutes, 40 cycles of denaturation at 95 °C for 15 seconds, annealing and elongation at 60 °C for 1 minute, and one cycle of denaturation at 95 °C for 15 seconds, annealing and elongation at 60 °C for 15 seconds, followed by denaturation at 95 °C for 15 seconds.

2.4 Data analyses and statistics

The SDS 2.3 software (Applied Biosystems) collected all results from RT-qPCR and the threshold was adjusted manually to 0.1. The dissociation and amplification curve for each amplicon was checked. Further analyses were processed in Microsoft Excel. The Pfaffl-method

(31)

29 was used to calculate the relative expression of the genes studied (Equation 1). This method takes into account the efficiency of the primers, in contrast to the ∆∆Ct-method, which assumes that all primers are 100% effective (Pfaffl, 2001). The geometric mean of the three housekeeping genes (ef-1-α, ẞ-actin and 18S rRNA) was used as reference genes in the Pfaffl- method to normalize experimental variation.

Equation 1:

𝑅𝑎𝑡𝑖𝑜 = (𝐸

_{𝑡𝑎𝑟𝑔𝑒𝑡}

)

∆𝐶𝑡 𝑡𝑎𝑟𝑔𝑒𝑡 (𝑐𝑜𝑛𝑡𝑟𝑜𝑙−𝑡𝑟𝑒𝑎𝑡𝑒𝑑)

(𝐸

_𝑟𝑒𝑓

)

∆𝐶𝑡 𝑟𝑒𝑓 (𝑐𝑜𝑛𝑡𝑟𝑜𝑙−𝑡𝑟𝑒𝑎𝑡𝑒𝑑)

Statistics were conducted in IBM SPSS statistics 23 and all graphs were made in GraphPad Prism (version 6.07 for Windows, GraphPad Prism Software, Inc). All data, both replicates and groups, were subjected to a normality test; Shapiro-Wilk test. Replicates showed that one of the two replicates in almost all groups failed to be normally distributed. Because of this, and the small replicate sizes (n=5), a non-parametric test called, Kolmogorov Smirnov test, was used to test if the duplicates within groups could be merged. The replicates were merged into their respective groups and thereby becoming a sample size of n=10. The groups (n=10) of all the genes at all time points were tested for normality by the same test as the replicates, showing that 91 % of the groups were normally distributed. Data that failed to be normality distributed showed no trends in other distributions (e.g. Bimodal), and no trend in skewness.

Since analysis of variance (ANOVA) has shown to be a robust test even with samples that have minor deviations from a normally distributed curve, and since the percentage of un-normal distributed groups were so small, one way ANOVA was used to check for significant differences between groups (Field, A. 2013). Games-Howell was used as a post-hoc test on groups that fails the Levine’s test of variance, and Gabriel’s procedure was used as a post-hoc test on groups that consisted Levine’s test of variance. The level of statistical significance for all tests was set at p< 0.05. An example from the statistical method conducted on the control group of one of the genes (crf1) are shown in Appendix III.

(32)

Results

30

3 Results

Expression levels of central HPI-axis genes (crf1, crf2, pomcA1, pomcA2, pomcB, gr1, gr2 and mr) in addition to two HSPs (hsp70a and hsp90a4) were examined in embryos and larvae of Atlantic salmon, subjected to handling; shocking and transport. The study was conducted to evaluate if stress during early development of fish alters the gene expression of the mentioned genes. In order to differentiate between the treatments and the group names, abbreviations were used for the three group names; shock (S), transport (T) and shock & transport (ST).

3.1 Hatching, mortality and larval growth

A hatching profile was made by counting the un-hatched eggs of all four groups, once a day during the hatching period (Figure 11). The results showed a small difference between the groups, where 50% of the eggs in group S were hatched 1-2 days before the other three groups. At the end of the hatching period, all groups showed a similar number of total hatched larvae.

5 0 0 5 1 0 5 2 0 5 3 0 5 4 0 5 5 0 5 6 0 5 7 0 5 8 0 0

1 0 2 0 3 0 4 0 5 0 6 0 7 0 8 0 9 0 1 0 0

H a t c h in g

d ° C

%

C S T S T

Figure 11: Hatching (%) of Atlantic salmon eggs during normal development, and after exposure to shock and/or transport.

Abbreviations for the four groups: C = control, S = shock, T = transport and ST = shock & transport.

(33)

Results

31 Dead eggs were removed from the hatching column once a day, and the amount was registered (Figure 12). Group S and ST showed a rapid increasing amount of dead eggs after shocking. Group T showed a rapid, but lower increase of dead eggs after transport. After hatching, the un-hatched eggs were removed, which led to a similar total amount of dead eggs in all four groups.

0 2 0 0 4 0 0 6 0 0

0 1 0 2 0 3 0 4 0 5 0

E g g m o r t a lit y

d ° C

%

C S T S T

S h o c k

T r a n s p o r t

U n - h a t c h e d e g g s r e m o v e d

Figure 12: Mortality (%) of Atlantic salmon eggs during normal development, and after exposure to shock and/or transport.

Abbreviations for the four groups: C = control, S = shock, T = transport and ST = shock & transport.

Yolk sac larvae was weighed at three time points; after hatching (583d°C), between hatching and start feeding (688°C) and at the time of start feeding (918d°C). The results showed a similar increase during development in all four groups (Figure 13).

(34)

Results

32

5 8 3 d ° C 6 8 8 d ° C 9 1 8 d ° C

0 1 0 0 2 0 0 3 0 0

Y o l k s a c la r v a e

Weight (g)

C S T S T

Figure 13: Weight (g) of Atlantic salmon larvaes at 583d°, 688 d°C and 918 d°C. Abbreviations for the four groups: C = control, S = shock, T = transport and ST = shock & transport. (n=28-30).

3.2 Ontogeny and long term treatment effects

Eggs and larvae collected previous to shocking (326d°C), transportation (377d°C) and stress challenge (918d°C; start feeding), in addition to newly hatched larvae (583d°C), were examined to study the ontogeny of different genes important in the stress response, and to examine possible long-term effect after exposure to shocking and/or transport. All genes assessed in the study (crf1, crf2, pomcA1, pomcA2, pomcB, gr1, gr2, mr, hsp70a and hsp90a4), were expressed at all studied developmental stages. The results of the examined genes will be shown in detail in the next sections.

3.2.1 Ontogeny of the HPI-axis genes and long term treatment effects

Samples from the control group (C) at the different time points were used to examine the normal ontogeny of the genes included in this study. The results showing the normal ontogeny of the genes are shown in the following figures; crf1 and crf2 (Figure 14), pomcA1, pomcA2 and pomcB (Figure 15), and gr1, gr2 and mr (Figure 16). All genes in the control group showed an increase during development, with significantly higher gene expression levels in larvae than in embryos. The two gr’s however, showed a significant decrease during the embryogenesis before an increase at the start of the larval period (Figure 16).

(35)

Results

33 Both paralogues, crf1 and crf2 (Figure 14), showed a similar expression pattern in the embryos followed by a significant increase after hatching. Further, crf1, in contrast to an increasing crf2, showed a significant decrease in larvae at the time of start feeding. The groups S, T and ST showed a similar profile as the control group, with a significant increase of both crf1 and crf2 after hatching, and a steady or increased expression in larvae at start feeding. When it comes to any long-term effect of the different treatments, a significantly lower expression of crf1 was detected in group T compared to the control group in newly hatched larvae. This difference, however, was also shown in the embryo prior to transport, of which T (untreated) was expected to be equal to the control group. Transcripts of crf2 in start feeding larvae were lower in group S and ST compared to the control group. The level of expression of Crf2 was lower compared to crf1 throughout the development.

Figure 14: Relative expression levels of genes crf1 and crf2 in eggs and larvae of Atlantic salmon prior to shocking (326 d°C), prior to transport (377d°C), at hatching (583d°C) and at start feeding (918d°C). Abbreviations for the four groups;

C=control, S=shock, T=transport, ST=shock & transport. The changes in relative gene expressions were measured by quantitative RT-PCR, normalized to the geometric mean of ef-1-α, ẞ-actin and 18s rRNA expression, and calibrated to the lowest expression of crf2 in the control group. The calibrator (value 1) is marked with a red star. Each column is presented as mean of 6-10 individuals ± SEM. Arrows indicate in which groups and at what day degrees (d°C) shocking and transport happened, and columns with pattern are groups that have been treated. Non-capital and capital letters indicate significant differences (p<0.05) within a group, or between groups at the same time point, respectively.

C S T S T C S T S T C S T S T C S T S T

0 2 4 6 8 1 0

c r f 2

Relative expression

a

B b

a

b

a b a

b c

d

a a

A c B c

a S h o c k b

T r a n s p o r t

*

3 2 6 d ° C 3 7 7 d ° C 5 8 3 d ° C 9 1 8 d ° C

C S T S T C S T S T C S T S T C S T S T

0 1 0 2 0 3 0

c r f 1

Relative expression

a

b A b

a A a

c c

c

a

b

a

b

B a

b

a

B b S h o c k

T r a n s p o r t

HPI-axis and heat shock protein (HSP) gene transcripts, and their responsiveness to stress in Atlantic salmon (Salmo salar L.) embryos and larvae